Respect to you for that, thanks Chris.

Zs

OMS,
A question or 2 please...

I'm curious that you rate XLPE cables with a starting temp of 80 deg C.
This then means that the connected accessories are rated at above 70 deg C.
This is not "always" the case, or do your designs make it "always" the case?
Thus in the event of future maintenance, it could be possible for A.N. Other to replace say a 90 deg C rated unit with a 70 deg C rated unit due to their not checking with the O & M manuals...

We are talking about the SWA as a CPC Paul - I'm suggesting that it will get to 80C under fault so the R2 is corrected from 20C to 80C in the derivation of Zs

I'm not sure glanding the armoured into a thermoplastic enclosure is much help for earthing

regards

OMS

-------------------------
Failure is always an option

Ah right, my misundstanding, when you gave the temps, I took these as the cable operating temps at the point of fault, rather than the final temps, sorry!

Great insight OMS

Thanks everyone for all the information. Earthing arrangements BTW are TN-S, wanted to confirm the distribution circuit complied with 543.1.3.

Handy table to have Bruce thanks. You too Zs, for going the extra mile, sounds like a handy book to have around any chance of the details?

Very much appreciated thanks again for all the input great place for info.

Steve

I'm not sure which book Zs has at the office, but you could try here for typical armour resistance values.

Regards

OMS

-------------------------
Failure is always an option

I did too - generally the armour starting temp is taken to be 10 degrees lower than the conductor operating temp (at least that's what table 54.4 seems to say).
- Andy.

OK - I guess some seeds of confusion sown by me - not unusual -

We are looking at the armour as a CPC and evaluating resistance and reactance under normal and fault conditions and also evaluating Zs.

We normally have data for this type of cable at 20C - which is fine for short circuit conditions, but not really helpful for earth faults as we are essentially trying to determine if a circuit will disconnect quickly and without damage - so we need to evaluate conditions for a hot, loaded installation that's in use to determine the highest likely value of Zs and from that disconnection time at the (depressed) fualt level

With a limit of 90C on conductor insulation, there is always going to be a temperature gradient across the insulation, bedding and fillers so the armour is likely to be about 10C less than the conductor temperature.

from here we typically have two or three scenarios.

1 - we can evaluate the fault at the conductor operating temperature in normal use - a function of the load, and where we estimate that downstream faults don't have enough magnitude to cause much heating in the cable - typically the kind of evaluation you might make for a feeder cable when considering a fault on a downstream final circuit protected by a low ampacity CPD. So calculate Zs for the distributor at operating temperature and use that to determine Zdb for the final circuit

2 - we can evaluate at the maximim permitted conductor operating temperature - typically the approach if you want to evaluate a fault on the distributor cable or sub main - essentially what table 54.4 tells you (ie a reduction of 10c from either 70C or 90C depending on insulation type)

3 - We may be using devices not listed in BS 7671 for protection (MCCB's, ACB,s Neozed type fuses or fuses designed for thyristor operation. At this stage we may well want to evaluate the cable beyond say 90 - typically we would do that at the average of the maximum permitted operating temperature and the final limiting temperature so for XLPE it would be at (80 +200)/2 = 140C. For the line conductors that would be at (90 + 250)/2 = 170C. So, you correct resistance of the line conductors to 170C, correct the armour resistance to 140C, add in the reactance and evaluate Zs from that - simples

Any clearer ?

Regards

OMS

-------------------------
Failure is always an option

And that, my friends, is far more eloquent than suggesting that in the event of a long term fault the cores might be gently frying so I've got me coat already. Blinding post OMS.

My source: I think it is called 'Electrical Installation Calculations' and is by B D Jenkins and (?Initials?) Coates.

Edited to remove topical 911 joke before this thread is archived for reference.

Zs

Edited: 20 July 2012 at 05:04 PM by Zs

- steady, Zs

Mark Coates is the chap

The most recent revision of the book Zs refers to is here

Regards

OMS

-------------------------
Failure is always an option

Marvellous text. Not bed time reading but a good read anyhow!

I have a fair collection of technical books including several copies of Mark Coates and Brian Jenkins book mentioned above. The latest edition published by the ECA is an improvement over the previous editions and is very useful for anyone engaged in design

At the top of the tree of installation text books is Paul Cook's Commentary on the IEE Wiring Regulations 17th Edition published by the IET. I have several editions but if I could only take one book to my desert island it would be this one.

If anyone wants resistance and impedance data on SWA cables if you PM your email address I will send you my latest paper on the subject FOC.

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/

OK, OMS, it seems you are taking the operating temp of the cable to be @ 90 Deg C, thus armour at 80 deg C at the point of the fault occurring?
i.e. you are working the Zs out for a fully loaded installation which will be the worst case due to increased conductor temp, both in the cores and the armour, increasing the resistivity of these items.
IF I am correct, then you are designing to run the cable at 90 Deg C, so at the point of the bolted fault this is your starting temperature for the cable.
Presumably then you would be designing into 90 Deg C rated terminations?

Have I understood correctly?

The reason I am "dragging" this out is that I have seen many "designs/installs", not all physical I may add, with XLPE cables where the design "rating" would run them at 90 Deg C and their termination points are only rated at 70 deg C.
It also seems something that is not greatly thought about amongst many general electrical contractors. Neither is the fact that now most cable is xlpe/swa/pvc, and that pvc/swa/pvc is, well at least in my area quite rare at the wholesalers!

However, I realise that under certain termination conditions even if the main body of the cable were running at 90 Deg C, the terminations may well run cooler, however, I suspect that this could be quite a transient situation.

Discuss?...

Regards

OMS

-------------------------
Failure is always an option

Thanks,

Paul

think of it this way Paul:

you have a radial circuit feeding say a few sockets fed from a DB with MCB protection.

That DB is fed in say 25mm2 XLPE armoured from a switchboard in the basement. That switchboard in turn gets it's supply via a distributor from say the energy centre that feeds a campus development (big hospital perhaps).

If you are evaluating the final circuit, Zs is the sum of the final circuit impedance, the sub main impedance, the distributor impedance. You probably want to calculate the impedance of the final circuit at 70C, the sub main at it's operating temperature and the distributor at ambient temperature with Ze just taken "as is".

For a fault on the sub main, you would calculate that at say 70C and the distributor at operating temperature.

For a fault on the distributor, you would probably evaluate that at 90C and add "something" for Ze (or rise in Transformer temperature)

For short circuit of course you want the conductors at 20c and would probably ignore much of the "resistance" anyway - just use reactance to give you a maximum fault level against which to design withstand ratings of kit

You can see that you need to calculate the sub main and distributor impedance several times at the relevant temperatures - which is a real pain in the ass if you are doing it long hand.

For "installed" installations then to be honest it's not worth the effort of going into the detail above - just bang the whole lot through at 70 or 90 c and the jobs a good un. You might not manage to reduce a cable size, but for boot laces then the cost variable isn't that high - labour outweighs it anyway.



It's a campus, Paul - 33kV intake, massive emergency and peak lopping on site generation and CHP - several hundred buildings - pretty impressive fault levels in places. The 50C rule applies everywhere though - so even in buildings that you could compare with "typical" DNO supplies, the combination of external fault level and run down effects of process machines does tend to make it "challenging" on occassion. The problem is normal sized cables tend to depress the fault quickly (ie over a short distance) - on that site though the cables are bigger to run cooler - so you get far less attenuation.

Regards

OMS

-------------------------
Failure is always an option

Thanks OMS, it does seem I understood what you were on about, even on the forum, that is unusual!

One last question, how much thought is put into these large installs for cable heating losses over the lifetime, I'm sure you are aware the the CDA push this and even have a free software programme for calculating the cost of cable heating losses & how much can be saved over its lifetime, admittedly at a fixed rate per unit.
Obviously the larger the cable, the less the Joule heating, the cooler it runs, the less risk there is of the resistivity taking you over the Zs, the greater the fault current, thus quicker disconnection.
Drawback is as you suggest, high fault currents carried a "long" way into the install.
I don't get involved with such heavily loaded install design to try to justify this to the client, and from a domestic standpoint it would be a non starter I suspect as the job cost would be the of the greatest concern rather than the whole life cycle costing.

I recall working at a now closed aluminium facility over in East Wales, before 2000, where by they were changing all motors to maximum efficiency when they were changed out as a cost saving exercise, fat lot of good that did them!

Mind perhaps this is too far off topic, sorry?

Again, discuss?...

Well cost of ownership and "standing" losses do play a part in design paul - either at inception/fesibility stage or for fine tuning designs.

as an example consider say a 1mVA load (hospital again) - you need redundancy so perhaps you have 2no 1MVA transformers (duty standby). First you would look at the % impedance - as that sets a ratio for no load to full load voltage and a limit on fault level - from there you decide if the balance of copper and iron losses warants running two transformers hot with a duty standby changeover (ie standby is unloaded and only seeingiron loss) or if you want to parallel them to give load sharing - each TX only now supplies 500kVA, the iron losses are the same but the copper losses drop in a square law relationship - so far less standing losses under base load - and instant recovery if you lose a single TX supply.

Moving on to the cable - well again, a hospital may have say 25% of peak load present at all times - so I2r losses in cables are worthy of a quick look - Amtech for example will give you cable losses at design current as part of the cable analysis tool.

Given that you also often need spare capacity on systems, it's quite easy to show a case for "whole life" cable sizing rather than a minimum to do a job - how viable that really is deponds on how complicated you make the payment model - but on a simple payback it's quite easy to show why a particular cable would be more economic over the installation life time (and it's an easy way to allow the designer to "up size" and reduce his design risk as well )

So does it play a part in mainstream design - yes, but only where there is some value in the process to warrant the extra effort. Domestically, I guess no (although teenage daughters, showers and 16mm2 might be worth further consideration !!)

Regards

OMS

-------------------------
Failure is always an option

I have looked up what i was trying to explain when i went off on a tangent lol.

If we take a cable Xple insulation, Pvc Sheath, operating temp 70 C, 25mm 4 core as above to BS 5476

The tables in the GN suggest a minimum of 36.1 and armour = 70mm steel

So table 54.7

k1/k2 x 16, So 70 is equivalent to 36.07

If i use the inital temp as 70C anf the final temp as 250C the K factor is 154.

So now the 70 mm equivalent is 48.31 yet the Xple can withstand a higher final temp.

I stand corrected to my blasé approach

So would i be right in saying the better approach would to assess the I2T as Zs suggested , and draw out an Adiabatic line for the armouring and a separate line for the live conductors.

Much appreciated.

Oh also if we take into account the corrected resistance and the magnetic effect of the armouring, is there a place to use the average of the initial and final temp in assessing I2T.

Again many thanks

regards

OMS

-------------------------
Failure is always an option